xref: /netbsd-src/external/bsd/nvi/dist/regex/WHATSNEW (revision dbd550ed1a6686d6600f748306f9cc03d8cd4c94)
1*dbd550edSchristos# @(#)WHATSNEW	8.3 (Berkeley) 3/18/94
2*dbd550edSchristos
3*dbd550edSchristosNew in alpha3.4:  The complex bug alluded to below has been fixed (in a
4*dbd550edSchristosslightly kludgey temporary way that may hurt efficiency a bit; this is
5*dbd550edSchristosanother "get it out the door for 4.4" release).  The tests at the end of
6*dbd550edSchristosthe tests file have accordingly been uncommented.  The primary sign of
7*dbd550edSchristosthe bug was that something like a?b matching ab matched b rather than ab.
8*dbd550edSchristos(The bug was essentially specific to this exact situation, else it would
9*dbd550edSchristoshave shown up earlier.)
10*dbd550edSchristos
11*dbd550edSchristosNew in alpha3.3:  The definition of word boundaries has been altered
12*dbd550edSchristosslightly, to more closely match the usual programming notion that "_"
13*dbd550edSchristosis an alphabetic.  Stuff used for pre-ANSI systems is now in a subdir,
14*dbd550edSchristosand the makefile no longer alludes to it in mysterious ways.  The
15*dbd550edSchristosmakefile has generally been cleaned up some.  Fixes have been made
16*dbd550edSchristos(again!) so that the regression test will run without -DREDEBUG, at
17*dbd550edSchristosthe cost of weaker checking.  A workaround for a bug in some folks'
18*dbd550edSchristos<assert.h> has been added.  And some more things have been added to
19*dbd550edSchristostests, including a couple right at the end which are commented out
20*dbd550edSchristosbecause the code currently flunks them (complex bug; fix coming).
21*dbd550edSchristosPlus the usual minor cleanup.
22*dbd550edSchristos
23*dbd550edSchristosNew in alpha3.2:  Assorted bits of cleanup and portability improvement
24*dbd550edSchristos(the development base is now a BSDI system using GCC instead of an ancient
25*dbd550edSchristosSun system, and the newer compiler exposed some glitches).  Fix for a
26*dbd550edSchristosserious bug that affected REs using many [] (including REG_ICASE REs
27*dbd550edSchristosbecause of the way they are implemented), *sometimes*, depending on
28*dbd550edSchristosmemory-allocation patterns.  The header-file prototypes no longer name
29*dbd550edSchristosthe parameters, avoiding possible name conflicts.  The possibility that
30*dbd550edSchristossome clot has defined CHAR_MIN as (say) `-128' instead of `(-128)' is
31*dbd550edSchristosnow handled gracefully.  "uchar" is no longer used as an internal type
32*dbd550edSchristosname (too many people have the same idea).  Still the same old lousy
33*dbd550edSchristosperformance, alas.
34*dbd550edSchristos
35*dbd550edSchristosNew in alpha3.1:  Basically nothing, this release is just a bookkeeping
36*dbd550edSchristosconvenience.  Stay tuned.
37*dbd550edSchristos
38*dbd550edSchristosNew in alpha3.0:  Performance is no better, alas, but some fixes have been
39*dbd550edSchristosmade and some functionality has been added.  (This is basically the "get
40*dbd550edSchristosit out the door in time for 4.4" release.)  One bug fix:  regfree() didn't
41*dbd550edSchristosfree the main internal structure (how embarrassing).  It is now possible
42*dbd550edSchristosto put NULs in either the RE or the target string, using (resp.) a new
43*dbd550edSchristosREG_PEND flag and the old REG_STARTEND flag.  The REG_NOSPEC flag to
44*dbd550edSchristosregcomp() makes all characters ordinary, so you can match a literal
45*dbd550edSchristosstring easily (this will become more useful when performance improves!).
46*dbd550edSchristosThere are now primitives to match beginnings and ends of words, although
47*dbd550edSchristosthe syntax is disgusting and so is the implementation.  The REG_ATOI
48*dbd550edSchristosdebugging interface has changed a bit.  And there has been considerable
49*dbd550edSchristosinternal cleanup of various kinds.
50*dbd550edSchristos
51*dbd550edSchristosNew in alpha2.3:  Split change list out of README, and moved flags notes
52*dbd550edSchristosinto Makefile.  Macro-ized the name of regex(7) in regex(3), since it has
53*dbd550edSchristosto change for 4.4BSD.  Cleanup work in engine.c, and some new regression
54*dbd550edSchristostests to catch tricky cases thereof.
55*dbd550edSchristos
56*dbd550edSchristosNew in alpha2.2:  Out-of-date manpages updated.  Regerror() acquires two
57*dbd550edSchristossmall extensions -- REG_ITOA and REG_ATOI -- which avoid debugging kludges
58*dbd550edSchristosin my own test program and might be useful to others for similar purposes.
59*dbd550edSchristosThe regression test will now compile (and run) without REDEBUG.  The
60*dbd550edSchristosBRE \$ bug is fixed.  Most uses of "uchar" are gone; it's all chars now.
61*dbd550edSchristosChar/uchar parameters are now written int/unsigned, to avoid possible
62*dbd550edSchristosportability problems with unpromoted parameters.  Some unsigned casts have
63*dbd550edSchristosbeen introduced to minimize portability problems with shifting into sign
64*dbd550edSchristosbits.
65*dbd550edSchristos
66*dbd550edSchristosNew in alpha2.1:  Lots of little stuff, cleanup and fixes.  The one big
67*dbd550edSchristosthing is that regex.h is now generated, using mkh, rather than being
68*dbd550edSchristossupplied in the distribution; due to circularities in dependencies,
69*dbd550edSchristosyou have to build regex.h explicitly by "make h".  The two known bugs
70*dbd550edSchristoshave been fixed (and the regression test now checks for them), as has a
71*dbd550edSchristosproblem with assertions not being suppressed in the absence of REDEBUG.
72*dbd550edSchristosNo performance work yet.
73*dbd550edSchristos
74*dbd550edSchristosNew in alpha2:  Backslash-anything is an ordinary character, not an
75*dbd550edSchristoserror (except, of course, for the handful of backslashed metacharacters
76*dbd550edSchristosin BREs), which should reduce script breakage.  The regression test
77*dbd550edSchristoschecks *where* null strings are supposed to match, and has generally
78*dbd550edSchristosbeen tightened up somewhat.  Small bug fixes in parameter passing (not
79*dbd550edSchristosharmful, but technically errors) and some other areas.  Debugging
80*dbd550edSchristosinvoked by defining REDEBUG rather than not defining NDEBUG.
81*dbd550edSchristos
82*dbd550edSchristosNew in alpha+3:  full prototyping for internal routines, using a little
83*dbd550edSchristoshelper program, mkh, which extracts prototypes given in stylized comments.
84*dbd550edSchristosMore minor cleanup.  Buglet fix:  it's CHAR_BIT, not CHAR_BITS.  Simple
85*dbd550edSchristospre-screening of input when a literal string is known to be part of the
86*dbd550edSchristosRE; this does wonders for performance.
87*dbd550edSchristos
88*dbd550edSchristosNew in alpha+2:  minor bits of cleanup.  Notably, the number "32" for the
89*dbd550edSchristosword width isn't hardwired into regexec.c any more, the public header
90*dbd550edSchristosfile prototypes the functions if __STDC__ is defined, and some small typos
91*dbd550edSchristosin the manpages have been fixed.
92*dbd550edSchristos
93*dbd550edSchristosNew in alpha+1:  improvements to the manual pages, and an important
94*dbd550edSchristosextension, the REG_STARTEND option to regexec().
95